Critical Care Suite on-device AI

The X-ray industry's first on-device AI suite for detecting and triaging critical conditions
At a Glance

Detect and triage critical conditions

Provides an immediate notification for the presence of pneumothorax*

Confidence at point of care

Provides an accurate and automated measurement of ETT position within seconds

High accuracy

Detects nearly all pneumothoraces with high accuracy and displays an overlay* of location

Seamless integration

AI algorithms automatically analyze images without routing to a server

On-device AI provides critical insights at the point of care

Technologists, radiologists, and physicians are all under tremendous pressure to manage an ever-increasing number of cases. Every minute counts when dealing with high-risk procedures such as endotracheal intubation1 and critical conditions like pneumothorax.

Critical Care Suite (CCS) is the X-ray industry’s first collection of on-device AI algorithms for automated measurements, case prioritization, and quality control. It integrates seamlessly with your existing X-ray workflow to help you deliver the highest quality care to patients.
FEATURES

Helping you respond fast without compromising diagnostic precision

Embedding Critical Care Suite on a X-ray device provides radiologists and technologists with a host of benefits, including faster results, case prioritization and better quality control of the images right at the patient’s bedside.

Endotracheal tube positioning

27.4% of intubations outside of the operating room result in mispositioned ET Tubes, which can lead to severe complications²⁻⁶. Critical Care Suite provides an accurate and automated measurement of ET Tube position on the X-ray device within seconds of image acquisition. 

Confidence at point of care – The on-device algorithm automatically detects the presence of ET tube in AP chest X-ray images. The average error is less than 1.0 mm when calculating the endotracheal tube tip-to-carina vertical distance⁷. The vertical distance between the tube tip and the carina is automatically calculated and displayed on device.

ET Tube cases at a glance – Enables immediate access to AI-derived measurements in PACS worklist via configurable DICOM® tags, and displays AI-generated measurements with an image overlay in PACS.

 

Pneumothorax detection and triage

Given the high number of chest X-rays ordered as “STAT,” i.e., with immediate priority, the triaging of true STATs has become challenging for bedside physicians and radiologists. An average hospital can miss 95 to 352 pneumothoraces each year8-10.

High accuracy11  – Partially localizes 100% of all detected large PTXs and 96.23% of all detected small PTXs.* Limits false alerts (94% specificity), with an Area Under the Curve (AUC) of 0.96.

Triage notifications  – Sends a secondary capture DICOM image to PACS and presents the AI results to the radiologist. Image flags help enable worklist prioritization and have the potential to expedite review of critical findings.

Overlay display* –  When a pneumothorax is detected, an overlay is displayed both on-device and in PACS to assist with localization, along with a graphical representation of the AI algorithm’s confidence level in detection.

Real-time quality alerts

Critical Care Suite includes Quality Care Suite, featuring AI algorithms that operate in parallel and help technologists reduce image quality errors and improve efficiency. 

Intelligent Auto Rotate – Automatically detects if an image is not upright and rotates the image to be upright for the viewer. It saves approximately 3–4 user interface clicks on 85% of mobile X-ray exams, saving approximately 187,000 clicks a year.12

Intelligent Protocol Check – Conducts an automated quality check to detect errors on the acquisition system, such as mismatch between the protocol used and the anatomy acquired.

Intelligent Field of View – Detects when a lung field is clipped in a frontal chest X-ray (AUC >0.99) and allows technologists to determine if a repeat is required before sending the image to PACS.13

Seamless integration with your X-ray imaging workflow

Critical Care Suite's on-device AI algorithms analyze X-rays without routing images to a server. The AI output is sent directly to PACS via secondary capture DICOM image.
Disclaimers:
*Feature only available on Critical Care Suite 2.1. Not commercially available in all markets. Check with your local GE HealthCare representative for availability in your country.
References
1. Stephen E. Lapinsky. Endotracheal intubation in the ICU. Crit Care. 2015; 19(1): 258.

2. World Health Organization Report -Communicating Radiation Risks in Pediatric Imaging.

3. Jemmett ME, Kendal KM, Fourre MW, Burton JH. Unrecognized misplacement of endotracheal tubes in a mixed urban to rural emergency medical services setting. Acad Emerg Med 2003;10:961–5. 

4. Katz SH, Falk JL. Misplaced endotracheal tubes by paramedics in an urban emergency medical services system. Ann Emerg Med 2001;37:32–7. 

5. Lotano R, Gerber D, Aseron C, Santarelli R, Pratter M. Utility of postintubation chest radiographs in the intensive care unit. Crit Care 2000;4:50–3. 

6. McGillicuddy DC et al. Is a postintubation chest radiograph necessary in the emergency department? Rachh, Pratik, et al. “Reducing STAT Portable Chest Radiograph Turnaround Times: Int J Emerg Med 2009;2:247–9.

7. Clinical study provided in K211161 clearance.

8. Yarmus, Lonny, and David Feller-Kopman. Pneumothorax in the critically ill patient. Chest 141.4 (2012): 1098-1105. 

9. Robert M Rodriguez et al. Pneumothorax and Hemothorax in the Era of Frequent Chest Computed Tomography for the Evaluation of Adult Patients With Blunt Trauma. doi:10.1016/j.annemergmed.2018.08.423. Epub 2018 Oct 2.

10. Julia Ley-Zaporozhan et al. Enhanced pneumothorax visualization in ICU patients using portable chest radiography. https://doi.org/10.1371/journal.pone.0209770.

11. GE HealthCare 510(k) K223491.

12. Based on anonymous usage data collected from 775 GE HealthCare X-Ray systems over a 6-month period ending in July 2022.

13. GE HealthCare 510(k) K183182.

Have a question? We would love to hear from you.

JB07687XE